First, make sure you ran vacuum analyze to update
the statistics for the table.
If a large portion of your table is going to be scanned,
Seq Scan is often faster than Index Scan due to possibly
random seeks within the heap file (the transaction commit
state isn't in the index, so there is still
Hi *,
I have pgsql7.0.2 on Linux2.2.16 and table with following indices:
CREATE TABLE T1 (NUM INT NOT NULL, NAME VARCHAR(10) NOT NULL, POP
VARCHAR(10) NOT NULL);
CREATE INDEX T1_I_NUM ON T1(NUM);
CREATE INDEX T1_I_NAME ON T1(NAME);
CREATE INDEX T1_I_POP ON T1(POP);
... and some data, but I real
